Finally cracked it... there was an error that only showed up on opening
the project about an ambiguous baseURI shared with two files. One of them
was a deeply nested import in the context graph. when I resolved this and
reopened the project the script produced outputs.
I think there needs to be a little more help given - if things are so
sensitive (and in this case to a TBC-specific implementation choice of
namespace/file correlation registration) then run-time should not just fail
without relevant information.
i would suggest that the ui:graphWithImports element should throw a
(meaningful and problematic resource identifying) error if there was some
previous error loading these graphs, because its not going to be visible to
the SWP user.
On Tuesday, 15 May 2018 16:17:59 UTC+10, Rob Atkinson wrote:
>
>
> ok - it works with the other file path (!!!!) but as you can see from the
> snippet - it is using the same file path to load the data,
>
> the other thing to note is that this is an external file project - not
> contained inside the workspace (IMHO all my projects need to be so i can
> test them against multiple versions of TBC if necessary)
>
> <ui:setContext ui:queryGraph=\"{= ui:graphWithImports(<
> http://surroundaustralia.com/app/crips/domain/>) }\">
> <sml:ImportXMLFile sml:ignoreDoctype="true" sm:outputVariable=\"xml\"
> sml:sourceFilePath*=\"/TQ_models/test0.1/A*gentModelsXML/{= ?fileName }\">
> <sml:ConvertXMLToRDF sml:baseURI=\"
> http://surroundaustralia.com/app/crips/tmp/am_sxml/\"; sml:replace=\"{=
> true }\" sml:xml=\"{= ?xml }\">
> <ui:update ui:updateQuery="{! INSERT { GRAPH ui:tempGraph { ?s ?p
> ?o } } WHERE { ?s ?p ?o } }" />
>
> </sml:ConvertXMLToRDF>
> </sml:ImportXMLFile>
> <ui:dumpGraph ui:graph="{= ui:tempGraph }" ui:filePath=
> *\"/TQ_models/test0.1/*tmp/agentmodels_as_sxml.ttl\" />
>
> </ui:setContext>
>
> If i run the test case using this output file it claims success but does
> not write the file!
>
>
>
> On Tuesday, 15 May 2018 16:02:40 UTC+10, Holger Knublauch wrote:
>>
>> From the source code, line 202 is failing if the file cannot be resolved
>> to a valid path. It's the same problem as before. Could you try it with the
>> same output file as the test case, to reduce the number of unknowns that I
>> cannot see?
>>
>> Holger
>>
>>
>> On 15/05/2018 15:54, Rob Atkinson wrote:
>>
>>
>> Well - i get a different error if the path root is invalid - so what is
>> "GuessLang" and how do I make it behave? (What is a "mismatch" in this
>> context?)
>>
>> test case indeed worked with SWP preferences on - but perhaps this vital
>> context ought to be visible in the help on SWP, and come up when you search
>> "TestCases" in help!
>>
>>
>> On Tuesday, 15 May 2018 15:37:28 UTC+10, Holger Knublauch wrote:
>>>
>>>
>>>
>>> On 15/05/2018 14:18, Rob Atkinson wrote:
>>>
>>> Great - this is the sort of context that would be really helpful linked
>>> from any doc fragment that mentions files !
>>>
>>> Anyway - no luck with the path yet:
>>>
>>> Caused by: java.lang.NullPointerException
>>> at org.topbraid.core.io.IO.guessLang(IO.java:202)
>>> at
>>> org.topbraid.sparqlmotion.EclipseWorkspace.save(EclipseWorkspace.java:114)
>>> at
>>> org.topbraid.spin.swp.engine.control.internal.DumpGraphControlElement.run(DumpGraphControlElement.java:35)
>>>
>>> using
>>>
>>> <ui:setContext ui:queryGraph=\"{= ui:graphWithImports(<
>>> http://surroundaustralia.com/app/crips/domain/>) }\">
>>> <sml:ImportXMLFile sml:ignoreDoctype="true"
>>> sm:outputVariable=\"xml\"
>>> sml:sourceFilePath=\"/TQ_models/test0.1/AgentModelsXML/{= ?fileName }\">
>>> <sml:ConvertXMLToRDF sml:baseURI=\"
>>> http://surroundaustralia.com/app/crips/tmp/am_sxml/\"; sml:replace=\"{=
>>> true }\" sml:xml=\"{= ?xml }\">
>>> <ui:update ui:updateQuery="{! INSERT { GRAPH ui:tempGraph { ?s
>>> ?p ?o } } WHERE { ?s ?p ?o } }" />
>>>
>>> </sml:ConvertXMLToRDF>
>>> </sml:ImportXMLFile>
>>> <ui:dumpGraph ui:graph="{= ui:tempGraph }"
>>> ui:filePath="/TQ_models/test0.1/agentmodels_as_sxml.ttl" />
>>>
>>> </ui:setContext>
>>>
>>> as you can see the project where the file comes from is the same as the
>>> target .....
>>>
>>>
>>> Hmm. I don't see a possible problem other than a mismatch of files. I
>>> would need to get a more complete picture to try to reproduce this on my
>>> machine...
>>>
>>>
>>> - so went back to try your example...
>>>
>>> and when i added the test case to the example.org project, then tried
>>> to access it according to this:
>>>
>>> Test Cases
>>> TopBraid Composer > SHACL Support
>>> <http://127.0.0.1:52621/help/ntopic/org.topbraidcomposer.core/html/SHACL.htm>
>>> Test Cases
>>>
>>> TopBraid Composer includes a capability to create and execute test cases
>>> that verify the correct definition of SHACL
>>> <http://127.0.0.1:52621/help/ntopic/org.topbraidcomposer.core/html/TestCases.htm?resultof=%22%74%65%73%74%63%61%73%65%22%20%22%74%65%73%74%63%61%73%22%20#SHACL.htm>
>>>
>>> shapes and functions. In order to use this feature, create a new SHACL file
>>> containing .test. in its file name. The easiest way to do that is to
>>> use the *File > New > RDF Test Cases File* wizard. Test files must
>>> import the DASH library, and may import other namespaces containing
>>> definitions you want to test.
>>> no test cases show up in the view when looking at this file...
>>> .. if i run it shows up in the results and claims success
>>> ... no file "out.ttl" appears in example.org project (even after
>>> refresh)
>>>
>>>
>>> I just verified it works with 5.5 but don't forget to activate Debug
>>> mode in Preferences > TopBraid Composer > SWP
>>>
>>> Holger
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> On 15 May 2018 at 14:02, Holger Knublauch <hol...@topquadrant.com>
>>> wrote:
>>>
>>>>
>>>>
>>>> On 15/05/2018 13:49, Rob Atkinson wrote:
>>>>
>>>> Thanks Holger,
>>>>
>>>> its difficult to keep track - but are sometimes path names are relative
>>>> to projects or are they always absolute? but i see the clue was in the
>>>> doc
>>>> /exampleProject!
>>>>
>>>> this raise a couple of other questions:
>>>> 1) in a production EDG environment where we want to ingest files
>>>> programmatically, without the TBC project environment being present, how
>>>> are such pathnames interpreted?
>>>>
>>>>
>>>> TBC is not needed for production but the Eclipse workspace will also be
>>>> used in an EDG/TBL environment.
>>>>
>>>> for example, I can see if we do a "one-shot" upload and process that
>>>> this works:
>>>>
>>>> _fileUpload If true then files uploaded as part of a multi-part HTTP
>>>> request will be copied into the workspace as temporary files. The name of
>>>> the corresponding file will be assigned to the argument variable. For
>>>> example, define a service with an argument arg:file of type string and
>>>> create a web form that uploads files as "file". The variable ?file will
>>>> then contain the path to the temp file in the workspace. You can then use
>>>> the function smf:baseURI to get the URI of that file, e.g. to open it
>>>> using <ui:setContext ui:queryGraph="..." />. This is particularly
>>>> useful for Excel files.
>>>> what if we wanted to retain the files - they are not "temporary" and we
>>>> may want to use them later, or if we, for example, wanted to go back and
>>>> try re-processing a file, how would we know how to find it - is there any
>>>> way for example to look at a directory and see the files.
>>>>
>>>>
>>>> If you don't use _fileUpload, XML files will become plain string values
>>>> that you can save to anywhere, e.g. using sml:ExportToXMLfile. Then use
>>>> the
>>>> magic property tops:files as in ?path tops:files ("/
>>>> myproject.org/files/xml" false) to iterate through a folder of your
>>>> choice.
>>>>
>>>>
>>>>
>>>> 2) is there a way to get the current project context, or look up a
>>>> project's local choice of path using some metadata, or set a base context
>>>> for file paths to be interpreted? We already have a situation where the
>>>> eclipse import for different people have a different root because some of
>>>> us are looking at a single project, others have opened a directory
>>>> containing multiple projects, and the scripts not longer work :-(
>>>>
>>>>
>>>> To be completely flexible, and to discover available files and folders
>>>> dynamically, use tops:files as above. Another approach would be to
>>>> parameterize the execution using some kind of config file, for example a
>>>> dedicated namespace that holds local settings, and that is expected to be
>>>> present in each installation.
>>>>
>>>> Holger
>>>>
>>>>
>>>>
>>>>
>>>> rob
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On 15 May 2018 at 13:19, Holger Knublauch <hol...@topquadrant.com>
>>>> wrote:
>>>>
>>>>> Hi Rob,
>>>>>
>>>>> ui:dumpGraph requires a graph URI, so the solutions with one of the
>>>>> temporary unnamed graphs would not work (as you have found out). However,
>>>>> using ui:tempGraph does work as confirmed in the attached test case.
>>>>>
>>>>> The error handling with ui:tempGraph is not ideal (and I have just
>>>>> added a better message for the next release), but what the null pointer
>>>>> means is that it cannot resolve the path in your workspace.
>>>>>
>>>>> In your case, /tmp/agentmodels_as_sxml.ttl, is there a project called
>>>>> "tmp" in your TopBraid/Eclipse workspace? It needs to, e.g. if you had a
>>>>> project "example.org" then use /example.org/file.ttl".
>>>>>
>>>>> If there is a different null pointer in your case, could you send me
>>>>> the stack trace?
>>>>>
>>>>> Thanks,
>>>>> Holger
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On 15/05/2018 10:56, Rob Atkinson wrote:
>>>>>
>>>>> OK - so I want to try using SPINMap to transform what default import
>>>>> looks like into the target ontology...
>>>>>
>>>>> so I'm trying to save the imported graph programmatically, because I'm
>>>>> still not sure if manual import gives the exact same graph as import
>>>>> using
>>>>> the sml:ImportXML - and AFAICT SPINMap needs a project file to load for
>>>>> it
>>>>> its UI based mapping processes
>>>>>
>>>>> - so, load and save example, set up SPINMap, invoke transformation
>>>>> using SWP for multiple files in the same XML schema...
>>>>>
>>>>> tried various things to dump the graph : (NB invoking
>>>>> with &_contextdebug=true)
>>>>>
>>>>> If none of these work, noting the graph loads (JSON print out example
>>>>> provided by Holger works), then what will work and what is wrong with
>>>>> these
>>>>> options?
>>>>>
>>>>>
>>>>> <ui:setContext ui:queryGraph=\"{= ui:graphWithImports(<
>>>>> http://surroundaustralia.com/app/crips/domain/>) }\">
>>>>> <sml:ImportXMLFile sml:ignoreDoctype="true"
>>>>> sm:outputVariable=\"xml\"
>>>>> sml:sourceFilePath=\"/TQ_models/test0.1/AgentModelsXML/{= ?fileName }\">
>>>>> <sml:ConvertXMLToRDF sml:baseURI=\
>>>>> "http://surroundaustralia.com/app/crips/tmp/am_sxml/\";
>>>>> <http://surroundaustralia.com/app/crips/tmp/am_sxml/%5C>
>>>>> sml:replace=\"{= true }\" sml:xml=\"{= ?xml }\">
>>>>> <ui:dumpGraph ui:graph=\
>>>>> "http://surroundaustralia.com/app/crips/tmp/am_sxml/\";
>>>>> <http://surroundaustralia.com/app/crips/tmp/am_sxml/%5C>
>>>>> ui:filePath="/tmp/agentmodels_as_sxml.ttl"
>>>>> />
>>>>> </sml:ConvertXMLToRDF>
>>>>> </sml:ImportXMLFile>
>>>>>
>>>>> runs with no error - but seems to dump nothing
>>>>>
>>>>> <ui:setContext ui:queryGraph=\"{= ui:graphWithImports(<
>>>>> http://surroundaustralia.com/app/crips/domain/>) }\">
>>>>> <sml:ImportXMLFile sml:ignoreDoctype="true"
>>>>> sm:outputVariable=\"xml\"
>>>>> sml:sourceFilePath=\"/TQ_models/test0.1/AgentModelsXML/{= ?fileName }\">
>>>>> <sml:ConvertXMLToRDF sml:baseURI=\
>>>>> "http://surroundaustralia.com/app/crips/tmp/am_sxml/\";
>>>>> <http://surroundaustralia.com/app/crips/tmp/am_sxml/%5C>
>>>>> sml:replace=\"{= true }\" sml:xml=\"{= ?xml }\">
>>>>> </sml:ConvertXMLToRDF>
>>>>> </sml:ImportXMLFile>
>>>>> <ui:dumpGraph ui:graph=\
>>>>> "http://surroundaustralia.com/app/crips/tmp/am_sxml/\";
>>>>> <http://surroundaustralia.com/app/crips/tmp/am_sxml/%5C>
>>>>> ui:filePath="/tmp/agentmodels_as_sxml.ttl"
>>>>> />
>>>>> runs with no error - but seems to dump nothing
>>>>>
>>>>>
>>>>> <ui:setContext ui:queryGraph=\"{= ui:graphWithImports(<
>>>>> http://surroundaustralia.com/app/crips/domain/>) }\">
>>>>> <sml:ImportXMLFile sml:ignoreDoctype="true"
>>>>> sm:outputVariable=\"xml\"
>>>>> sml:sourceFilePath=\"/TQ_models/test0.1/AgentModelsXML/{= ?fileName }\">
>>>>> <sml:ConvertXMLToRDF sml:baseURI=\
>>>>> "http://surroundaustralia.com/app/crips/tmp/am_sxml/\";
>>>>> <http://surroundaustralia.com/app/crips/tmp/am_sxml/%5C>
>>>>> sml:replace=\"{= true }\" sml:xml=\"{= ?xml }\">
>>>>> <ui:dumpGraph ui:graph="{= ui:tempGraph }"
>>>>> ui:filePath="/tmp/agentmodels_as_sxml.ttl" />
>>>>> </sml:ConvertXMLToRDF>
>>>>> </sml:ImportXMLFile>
>>>>>
>>>>> Gives a null pointer
>>>>>
>>>>> <ui:setContext ui:queryGraph=\"{= ui:graphWithImports(<
>>>>> http://surroundaustralia.com/app/crips/domain/>) }\">
>>>>> <sml:ImportXMLFile sml:ignoreDoctype="true"
>>>>> sm:outputVariable=\"xml\"
>>>>> sml:sourceFilePath=\"/TQ_models/test0.1/AgentModelsXML/{= ?fileName }\">
>>>>> <sml:ConvertXMLToRDF sml:baseURI=\
>>>>> "http://surroundaustralia.com/app/crips/tmp/am_sxml/\";
>>>>> <http://surroundaustralia.com/app/crips/tmp/am_sxml/%5C>
>>>>> sml:replace=\"{= true }\" sml:xml=\"{= ?xml }\">
>>>>> <ui:update ui:updateQuery="{! INSERT { GRAPH ui:tempGraph { ?s
>>>>> ?p ?o } } WHERE { ?s ?p ?o } }" />
>>>>> <ui:dumpGraph ui:graph="{= ui:tempGraph }"
>>>>> ui:filePath="/tmp/agentmodels_as_sxml.ttl" />
>>>>> </sml:ConvertXMLToRDF>
>>>>> </sml:ImportXMLFile>
>>>>>
>>>>> Gives a null pointer
>>>>>
>>>>> <ui:setContext ui:queryGraph=\"{= ui:graphWithImports(<
>>>>> http://surroundaustralia.com/app/crips/domain/>) }\">
>>>>> <sml:ImportXMLFile sml:ignoreDoctype="true"
>>>>> sm:outputVariable=\"xml\"
>>>>> sml:sourceFilePath=\"/TQ_models/test0.1/AgentModelsXML/{= ?fileName }\">
>>>>> <sml:ConvertXMLToRDF sml:baseURI=\
>>>>> "http://surroundaustralia.com/app/crips/tmp/am_sxml/\";
>>>>> <http://surroundaustralia.com/app/crips/tmp/am_sxml/%5C>
>>>>> sml:replace=\"{= true }\" sml:xml=\"{= ?xml }\">
>>>>>
>>>>> </sml:ConvertXMLToRDF>
>>>>> </sml:ImportXMLFile>
>>>>> <ui:dumpGraph ui:graph="{= ui:tempGraph }"
>>>>> ui:filePath="/tmp/agentmodels_as_sxml.ttl" />
>>>>> Gives a null pointer
>>>>> --
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "TopBraid Suite Users" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to topbraid-user...@googlegroups.com.
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>>
>>>>> --
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "TopBraid Suite Users" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to topbraid-user...@googlegroups.com.
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>
>>>> --
>>>> You received this message because you are subscribed to the Google
>>>> Groups "TopBraid Suite Users" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>> an email to topbraid-user...@googlegroups.com.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>>
>>>> --
>>>> You received this message because you are subscribed to the Google
>>>> Groups "TopBraid Suite Users" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>> an email to topbraid-user...@googlegroups.com.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "TopBraid Suite Users" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to topbraid-user...@googlegroups.com.
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>>
>>> --
>> You received this message because you are subscribed to the Google Groups
>> "TopBraid Suite Users" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to topbraid-user...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
--
You received this message because you are subscribed to the Google Groups
"TopBraid Suite Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to topbraid-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
